Due to the rapid increase of the world population, animal productivity and feed security are becoming major challenges. The demand in feed supply is considerable, and more than 70% of cereal crops are dedicated to animal production. Therefore, any factor that affects the security of the feed supply is a significant constraint to production. Feed spoilage by fungi is not a new problem, but due to their great adaptability these microorganisms are posing a serious risk to the animal feed industry. Recently, fungi have been designated as a greater threat to animal, plant and ecosystem health than the other taxonomic classes of pathogens [1]. In addition, the affected commodity may become contaminated with toxic secondary fungal metabolites, known as mycotoxins. Mycotoxins are structurally diverse low-molecular weight metabolites produced by various molds belonging chiefly to species of the Aspergillus, Penicillium and Fusarium genera. These toxins inflict loss to farmers and reduce the value of contaminated feeds. For example, in the USA, the economic cost due to three mycotoxins (aflatoxins, fumonisins and deoxynivalenol) is estimated to be USD 900 million per year [2]. Effects in animals following the ingestion of these fungal compounds vary from acute, overt disease with high morbidity and death to chronic, decreased resistance to pathogens and reduced animal productivity [3,4]. However, the major problem associated with animal feed contaminated with mycotoxins is not acute disease episodes, but rather the ingestion of low level of toxins which may cause an array of metabolic, physiologic, and immunologic disturbances [3,4].

Although the literature about mycotoxins is rich in reports investigating cellular mechanisms, cellular toxicity, associated pathology and animal performance, studies on the effect of these compounds on the gastrointestinal tract (GIT) is limited. Considering the initial interaction of mycotoxins is with the gut epithelium, this topic has gained significant research interest in the last decade for a number of reasons. First, most acknowledge that a healthy intestinal tract guarantees the welfare and the health of both people and animals. Second, rapidly dividing and activated cells and tissues with a high protein turnover are predominant in gut epithelium. Intestinal cells and tissues can become a main target of mycotoxins as many of these metabolites are inhibitors of protein synthesis. Third, the absorption of mycotoxins and their fate within the GIT suggests that the epithelium is repeatedly exposed to these toxics, and at higher concentrations than other tissues. This latter point is further explored in the next section of this review.

The maintenance of a healthy GIT is crucial as it insures that nutrients are absorbed at an optimum rate, it provides efficient protection against pathogens through its own immune system, and it maintains the indigenous microflora in adequate numbers and confined to their natural niches. These three functions of the GIT might be pictured such as a “ménage à trois” in which each component interacts with each other (nutrition ↔ immune system ↔ gut microbiota ↔ nutrition) to maintain intestinal homeostasis. Whenever the integrity of the intestinal mucosa is compromised, nutrient absorption decreases. In addition, an increased proportion of absorbed nutrients is directed to repair the damaged area and to support the immune system until the intestinal insult is eliminated [5].

If one component from the “ménage à trois” is compromised, the adverse effect might be extended to the other components. For example, altering the numbers and species of GIT bacteria may affect the host’s ability to digest food and to stimulate the immune system. In this review we have therefore summarized recent findings following mycotoxin exposure on digestive and absorptive functions, intestinal defense, and microbiome composition. This review is focused on the major mycotoxins in terms of occurrence and toxicity, namely aflatoxins (AF), ochratoxin A (OTA), deoxynivalenol (DON), T-2 toxin (T-2), zearalenone (ZEA) and fumonisins (FB). Seven intestinal processes were investigated as shown in the Table 1. In addition, animal responses were evaluated according to the experimental doses used. Based on recent mycotoxin surveys [6,7,8], these doses were placed in three different categories: realistic, occasional and unrealistic doses (Table 2). This approach was applied to all results reported in this review.

Mycotoxin uptake and subsequent tissue distribution is governed by GIT absorption. This passage across the intestinal barrier may be maximal, as with aflatoxins (AF), or very limited, as with fumonisins (FB) (Figure 1). The bioavailability of these fungal compounds is thereby very diverse and differs between animal species (Figure 1). Regardless, the intestinal epithelium is exposed to the entire content of contaminated feed and is the first target of these contaminants. The rapid appearance of most mycotoxins in the circulation clearly indicates that the majority of the ingested toxin is absorbed in the proximal part of the GIT [9,10]. Mycotoxins can therefore compromise the intestinal epithelium either before absorption in the upper part or throughout the entire intestine by non-absorbed toxins. Indeed, with the exception of AF which is absorbed at high rates regardless of the species [10], absorption of other mycotoxins, such as trichothecenes (TCT), OTA, or FB may vary from 1% to 60% [9,11,12,13]. Thus, a substantial portion of non-absorbed toxin remains within the lumen of the GIT. The poor intestinal absorption of FB, ranging from 1% to 6% in non-ruminant species [11], implies that gut epithelium is exposed to a very high proportion of the toxin ingested. Similarly, absorption of deoxynivalenol (DON) is moderate in pigs, but very limited in poultry (Figure 1) [13,14]. The relative tolerance of poultry to DON has been partly attributed to its low bioavailability. However, the potential impact of the remaining DON in the intestinal lumen is still unknown, and the tolerance level within the GIT might be different. However, in comparison to pigs, the intestinal transit time is very rapid in poultry, and therefore this may reduce the exposure time to this mycotoxin. More importantly, several mycotoxins have been shown to undergo entero-hepatic circulation (Figure 1) [9,12,15,16].

This makes the mycotoxins available again via the bile in the entero-hepatic cycle, resulting in reabsorption and a prolonged retention time in the GIT. Entero-hepatic recycling must contribute to FB toxicity as intestinal absorption of this toxin is very low. Thus, in the intestinal lumen FB could be incorporated into mixed micelles, through interactions with cholesterol and/or bile salts, and thereby facilitate its intestinal absorption [15]. Interestingly, an in vitro study showed that FB was unable to cross the intestinal epithelium [89], but considering this mode fails to mimic in vivo conditions (e.g., through entero-hepatic recycling or through incorporation into mixed micelles) conclusions must be drawn carefully. In addition to the increased exposure through entero-hepatic circulation, it has been recently suggested that DON first enters the blood circulation when absorbed in the upper GIT and then reenters the intestinal lumen, passing through the more distal intestinal cells from the blood stream via the basolateral side of the cell [50]. Although an in vitro study, this may explain why some effects of DON are observed in the mid and distal jejunum rather than the proximal intestine.

Mycotoxin metabolism can occur in both the liver and the digestive tract. Intestinal metabolism, whether be in the gut epithelium or by gut microorganisms, may limit the toxic effects of mycotoxins within the GIT. This is especially true for ruminants which are able to convert many mycotoxins into non-toxic metabolites. Ruminant resistance to some mycotoxins has been attributed to the detoxifying role of the microbial population in the rumen. Many compounds, such as ZEA, DON or OTA are effectively rendered non-toxic by rumen microorganisms before absorption [9]. However in non-ruminants, intestinal biotransformation of mycotoxins takes place predominantly in the large intestine and thus provides little detoxification prior to absorption. Unlike other mycotoxins, AF requires metabolism to its toxic metabolite. While this activation has been largely described in the liver, AFB₁ metabolism to the reactive epoxide also takes place in the intestinal tract [88].

The absorption and fate of mycotoxins within the GIT provides evidence that the intestinal epithelium is prone to the toxic action of these toxins. Therefore, this paper reviews the consequences of mycotoxin exposure on the different processes governing the major functions of the intestine.

It is very well documented that the major mycotoxins are able to adversely affect the growth of animals. Impact on performance varies according to many factors, such as the mycotoxin and the species used, the concentration in feed, the use of purified versus naturally contaminated feed, or the ingestion of multi-toxin contaminated feed. Although the effect of low doses is more controversial, the reduced performance is among the main characterized effects of mycotoxin intoxication [3]. Though decreased body weight gain of animals is seemingly a consequence of reduced feed intake or outright feed refusal, a significant body of research points to a direct and/or indirect effect of mycotoxins on the nutrient quality, digestibility and/or absorption. There is a close association between production performance and digestive activity. For example, authors from the 1970’s and 1980’s reported reduced absorption of essential nutrients after aflatoxin [3] and trichothecene (TCT) [36,64] exposure. A reduction in the severity of acute aflatoxicosis in broilers oversupplemented with energy has been noted [90]. Thus, it is important to elucidate how mycotoxins modulate the activity of enzymes and transporters involved in nutrient digestion and uptake, and subsequently, the consequences on nutrient digestibility and on metabolizable energy.

The gastrointestinal tract (GIT) possesses its own immune system and it is estimated that up to 70% of the immune defenses of the organism are located in the intestine. In the GIT, the mature immune system of animals consists in specific tissues, such as gut-associated lymphoid tissues (GALT; Peyer’s patch, mesenteric lymph nodes, cecal tonsils) where immunocompetent cells are able to mount an efficient immune response. Complementary to that, early and immediate responses are provided locally along the length of intestine, where mucus, intraepithelial immune cells as well as intestinal epithelial cells (IECs) play a key role as sentinels and defenders.

Though it is well accepted that mycotoxins are able to modulate immune responses [4], the consequence of this immunomodulation for the GIT has been less documented. The capacity of animals intoxicated with mycotoxins to regulate the intestinal immune balance and/or mount an appropriate intestinal immune response is addressed here.

Intestinal epithelial cells (IECs) have two crucial but conflicting processes. On one hand, they transport nutrients and fluids and, on the other hand, they restrict the access for luminal antigens to the internal milieu. They form a monolayer that constitutes a dynamic and selective barrier, and mediates the transport of molecules in two ways: either across the cells (i.e., the transcellular pathway) or between the cells (i.e., the paracellular pathway) (Figure 2). This polarized monolayer effectively separates the apical (luminal) from the basolateral compartment, i.e., the lamina propria. Tight junctions between adjacent cells represent an integral part of this compartmentalization and any damage to them leads to an enhanced permeability of the cell layer and a decreased transepithelial electrical resistance (TEER) which can lead to intestinal disorders.

Since 2000, many authors have focused on the effect of mycotoxins on intestinal permeability. Modulation of the intestinal barrier was mostly studied with in vitro models (Table 5). Once differentiated and formed into a polarized monolayer, IECs become a very useful tool. Measurement of TEER is then feasible, and is considered as a good indicator of the integrity of epithelial barrier. At different concentrations, mycotoxins, particularly DON, are able to significantly reduce the TEER (Table 5). This decrease could indicate an alteration in paracellular permeability, but this ion movement across the monolayer can, however, be caused by changes in transcellular ion flux through altered plasma membrane channels or pumps. Accordingly, to eliminate this possibility, some authors have evaluated the apical to basolateral flux of paracellular markers, such as dextrans or mannitol, and noticed an increased flux after either DON or OTA exposure (Table 5) [30,54,55,56]. Effects in the paracellular pathway suggest an adverse effect on tight junctions (TJ). These are multiprotein complexes that link adjacent epithelial cells near their apical border (Figure 2).

Proteins present in the TJ complexes include ZO-1, occludin, and one or more claudin isoforms. TJ seal the luminal end of the intercellular space and limit transport by this paracellular route to relatively small hydrophilic molecules. ZO-1 acts as a scaffold to organize transmembrane TJ proteins and recruits various signaling molecules to the complex. Occludin binds to ZO-1 and the actin cytoskeleton and appears to have a role in regulating permeability through the TJ [30]. However, numerous studies have pointed to the claudin family of TJ proteins as a key determinant of paracellular characteristics. These proteins appear to form the backbone of the TJ. Therefore, it was of interest to explore the effect of these fungal metabolites on the TJ network, especially on claudins, and to relate these findings to the reduced TEER and the increased permeability to markers previously observed by several authors. Either by immunofluorescence or immunoblotting, it appears that DON and OTA removed or reduced the expression of claudin 4 and 3 on IECs (Table 5) [30,31,50,52,54,55,56]. Importantly, the finding on claudin 4 was demonstrated in vivo in the jejunum of pigs fed low concentrations of DON for five weeks [54]. Since DON exposure resulted in a reduction in total protein synthesis [55], the reduced expression of claudin 4 was attributed to this phenomenon rather than increased degradation or delocalization. Indeed, Van de Walle et al. [55] noted that claudin 4 expression was not restored after using an inhibitor of protein degradation, and Pinton et al. [54] did not observe delocalization of this protein from IECs plasma membrane. Further, both authors failed to show an effect at the mRNA level. In line with that, Lambert et al. [31] suggested that the delivery of de novo claudins 3 and 4 to the TJ complex might be perturbed by mycotoxins and would result in a reduction of total cellular levels of these claudins. As a consequence, new molecules would not arrive at the TJ to replace molecules which have been turned over.

Oxidative stress could also participate in the effect of OTA on intestinal permeability [29,31]. Although FB₁ has been shown to reduce the TEER in IECs [73], no data on the claudins is available. However, because of the well known effect of FB₁ on sphingolipid metabolism [98], and the major role played by sphingolipids and lipid rafts in the establishment and maintenance of TJ [31], the gut epithelium might be prone to the adverse effects of FB₁ on the barrier function. In accordance with this, Bracarense et al. [51] observed a defective expression of occludin and E-cadherin in the ileum of piglets fed low doses of FB₁. In addition to the TJ complex, E-cadherins also play an important role in cell adhesion (Figure 2). With regard to occludin and ZO-1, inconsistent results have been reported (Table 5). Some authors did not report any effects of mycotoxins on the two proteins unlike claudins in their studies [30,54,55]. On the other hand, some authors showed a reduced expression of occludin or ZO-1 which was associated or not with an effect on claudins [50,51,57]. Interestingly, Diesing et al. [50] observed changes in TEER and in TJ proteins (ZO-1 and claudin 3) only when DON was applied on the basolateral side of IECs. The authors failed to show any effects of the mycotoxin after apical exposure, unlike other studies. This work provides a new insight into the effects of DON as there is evidence of the existence of an active DON transport in the basolateral to apical direction as opposed to simple diffusion from apical to basolateral in IECs [50]. Moreover, the application route of DON may explain why some authors failed to report effects on ZO-1 after apical incubation (Table 5). Similarly, change in occludin expression has been only reported in vivo, and not after in vitro apical exposure.

Further work has been done to elucidate the initial cellular mechanism leading to this disruption of the intestinal barrier. TJ structure and function can be regulated by signaling molecules involved in MAPK pathways, and DON is known to rapidly activate MAPK [99]. Based on this observation, it has been demonstrated that DON decreases the intestinal barrier function through a MAPK dependent mechanism. Indeed, DON-activated MAPK led to a decrease in claudin expression [52,56], and inhibition of ERK1/2 phosporylation with a specific MAPK inhibitor restored the barrier function of IECs [56].

Collectively, these data suggest that some mycotoxins, especially DON have the ability to increase intestinal permeability, allowing the entry of luminal antigens normally restricted to the gut lumen. Translocation of bacteria across IEC monolayers have been already reported after DON exposure [53,54], and this event is considered as a key step in the induction and persistence of inflammatory bowel diseases [97]. Nonetheless, further studies with animals are required considering most of the data were from in vitro studies. Even if cellular models delineate the mechanism of action of the toxins, it is important to demonstrate that the same effect/mechanism can be observed on primary tissues. In this regard, Pinton and co-workers assessed the intestinal toxicity of mycotoxins in vitro, using IEC line, ex vivo, using intestinal explants and in vivo, using intestinal tissues from animals exposed to mycotoxin contaminated diets [52,54]. It can be concluded that mycotoxins could promote intestinal disorders, and coupled with the previous findings on Na⁺-dependent glucose transport, could be the underlying cause of diarrhea in animals exposed to mycotoxins. More importantly, mycotoxins may also facilitate their own intestinal absorption through the paracellular pathway. Some authors reported an increased permeability to compounds up to 10-kDa [30,54] whereas the molecular mass of mycotoxins is usually less than 1-kDa. Furthermore, mycotoxins that are poorly absorbed in the intestine, such as FB, would reach the systemic circulation easier if the intestinal barrier is compromised. In line with that, ingestion of co-contaminated feed with DON and FB resulted in more pronounced effects than the ingestion of the mono-contaminated feed with FB [51,100]. However, the authors did not measure the content of this mycotoxin in the plasma of animals fed the contaminated diets. Similarly, it has been demonstrated that 15-ADON, an acetylated derivative of DON commonly produced together with DON, induced a greater impairment of the barrier function (as measured through TEER, TJ, MAPK) than DON [52]. These findings emphasize the need to assess the effect of feed contaminated with more than one mycotoxin.

In every species, intestinal microflora are an important factor for animal health because there are close links between the host and its’ intestinal microflora especially through immune responses and via the metabolic products of microbial fermentation. Thus an impaired balance of the intestinal microbiome, such as in a dysbiosis condition, could have many adverse effects on the health of the host. However, investigating the microbial community shift is still a complex and imprecise activity, partly due to the low culturability of many bacterial species from the gastrointestinal tract (which can vary from 10% to 50%) and the inability of classical bacteriological counts to illustrate the changes in individual species abundance of the microbial community. Although culture-independent methods, such as high-throughput sequencing platforms have been developed to overcome culturing biases, data on the influence of toxins on the intestinal microflora is still limited. These issues account for the lack of information of mycotoxin effect on the microbiome, in contrast to the data on eukaryotic cells. Most of the data available regarding mycotoxin interactions with the animal’s microbiome, concern the role of intestinal microfora in mycotoxin detoxification [101].

Mycotoxins not only undergo microbial metabolism in the rumen or intestine, but may affect the microbes and their communities as some toxins exhibit antimicrobial properties [16]. Such antimicrobial activities of mycotoxins are suspected of (i) affecting the fermentative capacity of the rumen [16] or (ii) favoring a shift toward intestinal aerobic bacteria such as observed in inflammatory bowel diseases [97]. The tolerance of ruminants to some mycotoxins is attributed to toxin metabolism by rumen microorganisms, the detoxifying potential of the rumen may be influenced by diet composition. Indeed, the intake of a diet low in structural and high in rapidly degradable carbohydrates will decrease rumen pH, and an acid pH has been demonstrated to inhibit the complete transformation of DON to its metabolites [47]. Using different concentrate proportions in ruminant rations, realistic doses of DON decreased the fermentation of fiber fractions at the lower pH value [47,48] indicating a restriction of cellulolytic microbes. The change in the microbial community of the genus Clostridium, which contains cellulolytic species, after inclusion of DON confirmed these findings [47]. As reviewed by Maresca and Fantini [97], mycotoxins might be potential risk factors for chronic intestinal inflammatory diseases. In line with that, the total number and composition of intestinal microflora are significantly modified in inflammatory bowel diseases, with an increase in the number of aerobic bacteria and a parallel decrease in the number of anaerobic bacteria. Feeding pigs with T-2 toxin resulted in a substantial increase of aerobic bacterial counts in the intestine [65]. Similarly, chronic exposure of pigs to low doses of DON caused an increase in the number of intestinal aerobic bacteria and modified the dynamics of the intestinal bacteria communities [49]. By contrast, the mycotoxin FB₁ did not alter the in vitro growth of isolated bacteria representative of intestinal microflora [72].

Mycotoxin research into effects on intestinal functions has made substantial progress in recent years By contrast to the limited distribution of mycotoxins into systemic tissues, the GIT is exposed to all the mycotoxins in contaminated feed. This suggests that the intestinal epithelium is the major site for the effects of mycotoxin contaminated material, even low levels of contamination. The influence of mycotoxins on intestinal balance is observed at relatively low levels that are not associated with obvious adverse effects on growth. Collectively, the data from research studies with realistic doses show that mycotoxins, and in particular DON, can compromise several intestinal functions, such as digestion, absorption, permeability, defense, and result in lower productivity and poor health of animals. In the future, much attention should be paid to low concentrations of mycotoxins, even though moderate doses can be encountered occasionally during unfavorable weather conditions. Besides, the consequences on physiological processes might be very different from those observed with high doses. Indeed, in this review we showed that low doses of mycotoxins are able to upregulate cytokine expression. By contrast, higher doses would lead to an opposite profile by downregulating them. Applegate et al., [19] reported a similar effect on sialic acid excretion and the activity of intestinal maltase when using increasing concentrations of AF. The authors suggested that these physiological responses were following a pattern of hormesis. Hormesis is a dose-response phenomenon characterized by low-dose stimulation and high-dose inhibition. Hormesis has been noted in regards to changes in body weight of chickens receiving graded levels of dietary AF [102].

As shown in the Table 1, experiments elucidating the effects of mycotoxins other than DON on the GIT are rather limited. Similarly, more studies should investigate the effects of diets contaminated by more than one mycotoxin. Most fungi are able to produce several mycotoxins simultaneously and the mycotoxins produced depends on the feedstuff and crop growing conditions [3] and this has been demonstrated in worldwide mycotoxin surveys [6]. As it is a common practice to use multiple grain sources in animal diets, the risk of exposure to several mycotoxins increases with diet complexity [91]. Authors should experiment with naturally contaminated feed as feeding naturally contaminated grains take into account the presence of masked mycotoxins and their precursors [103,104] as well as unidentified fungal metabolites that may contribute to an underestimation of the total amount of mycotoxins. Sometimes, these factors can make interpretation of results difficult. Moreover, the nutritive value of grains may be lower due to fungi invasion, and thereby may cause a greater effect on animal productivity [3]. In a future where climate change may significantly affect the worldwide distribution and contamination by mycotoxigenic fungi and mycotoxins [105], the analysis of levels of contamination as well as the implementation of prevention and control strategies will be of major concern.